黄土潜水层深井降水计算模型及黑河少陵塬隧洞的降水分析

【作者】 康婷；

【导师】 樊正复； 郭增玉；

【作者基本信息】 西安理工大学 ， 应用数学， 2001， 硕士

【摘要】 随着高层建筑、地铁、长大隧道等工程的发展。降低地下水位的问题日益突出。目前对降水方案合理性的检验大多采用经验公式或根据泰斯的平面流理论进行计算，与实际情况差距较大。本文对黄土地区潜水含水层深井降水问题作了较系统地研究，初步得到如下研究成果。 通过对少陵源区大量的抽水试验资料的整理，揭示了少陵塬区黄土厚潜水层深井抽水时的水流运动规律。经过分析发现在一定的抽水条件下，某点的潜水水位降深和此点含水层厚度的乘积与抽水时间呈对数关系变化。分析了水跃值的变化规律，当井出水量很大时，水跃值在短时间内达到最大值。而后随着抽水时间的延长缓慢减小；当井出水量较小时，在很长一段时间内，水跃值逐渐增大：井储容量对水跃值的变化影响很大。根据抽水试验资料建立了修正泰斯模型以及参数确定方法，将计算结果与试验资料对比，并对少陵源区的降水方案分析评价。表明此经验公式可以在黄土地区类似降水工程中应用。 黄土地区群井抽水时，井周土层中水流为明显的三维流，随着抽水时间的延长，潜水含水层的厚度逐渐减小。应用泰斯公式计算时，按照平面的二维流来处理，潜水含水层的厚度也取为常数。本文根据黄土的渗透特性以及深井抽水时的水流规律，建立了黄土厚潜水含水层深井抽水时三维水流模型。 应用C++语言编制了三维有限元程序。三维问题中，抽水井的处理是一个难点。在以前的二维程序中将抽水井简化为一个点，井出水量直接作用在这个点上，这种处理方法极易引起该点的水位异常；目前在三维问题中，由于抽水井穿越许多单元，此问题依然没有解决。本文试图将抽水井当作一系列虚单元来处理。此程序可以模拟单井抽水时计算区域各点水位降深随时间的变化，研究水流运动规律；求解任意边界条件下的群井定流量抽水问题，预测潜水位，指导施工开挖进度，评价降水方案的合理性。利用此程序计算了少陵塬区的降水问题，并与试验资料进行对比，验证了所建立的模型比较符合黄土地区潜水深井抽水时的水流运动规律，所编制的程序是正确的。更多还原

【Abstract】 With the development of the economy, constructions for large buildings with deepbasements,subways, tunnels,and many other structures often require the excavation ofsoil and rock to substantial depth below the natural water table. The problems aboutdewatering become more and more important. Now, the feasibility of dewatering planis tested by experiential formulas or TaiSi’s two dimensional law, thus big errors maybe resulted. This paper studied systematically the problems about dewatering throughdeep well pumping in thick phreatic water of loess. Some preliminary results arefollowing.I arranged data with well pumping tests in Shao Ling Yuan and sum up the laws offlows under the condition of deep well pumping in thick phreatic formations loess,which showed that, under a certain condition of pumping, the product of water table ina dot and the thickness of aquifer in this dot vary with pmping time by logarithm. Ianalyzed the law of well loss, if the rate of pumping is very large, water spring mountsto maximum in a short time,then with pumping extending well loss becoms smallslowly, if the rate of pumping is small, well loss becomes big gradually in a long time.The reserves of water in pumping well influence the change of well loss considerably.On the basis of experimental data, I set up corrective TaiSi’s model and studied themethod of parameters settlement in the model. The comparison with the calculatedresults and experimental data showed that the model is feasible and the method ofparameters settlement is correct. The experiential formula can be used into similardewatering projects in the other loess region.Pumping with group wells in loess, flows in formations surrounding group wells arethree dimensional obviously. With time of pumping, the thickness of phreaticformations become smaller and smaller. But when we calculate water table usingTaiSi’s formula, we consider flows as two dimensional flows, the thickness as constant.Considering above facts, I set up three dimensional flows model through deep wellpumping in thick phreatic formations of loess in this paper.I made three dimensional program with computer language C++, in which thehandle of pumping wells were very difficult. In previous two dimensional programs,wells were simplified a dot. The rate of pumping was considered that it was pumped outof this dot. According to my practice, water table calculated with this method is oftenextraordinary. Now in three dimensional programs, this problem has not been answeredyet because of wells cutting across many elements. In this paper, I attempt to considerthe group wells as many empty elemats. Using this program, we can calculate watertable every place in computing region, study the law of fiows, answer the problemsunder group wells pumping in every boundary condition, predict water table, gulde theprocedure of eation, evaluate the feasibility of dewatering plan, and revise thedesigned dewatering program. I calculatd the problems of Shao Ling Yuan with thisprogram and compare the calculated results with the experiential results. from whichshowed that the model accord with the law of flows under the condition of deep wellpamping in phreatic formations of loess and the program are correct.更多还原